fix: use cached user list in /check instead of per-member GetUserInfo calls

[Copilot]

handleListMissing was calling api.GetUserInfo(uid) once per team member on every /check invocation — N serial Slack API calls adding latency and rate-limit exposure proportional to team size.

Changes

• handleListMissing: Replace the per-member api.GetUserInfo loop with a single getCachedUsers(api) call (5-minute TTL cache) before the loop, then build an id→slack.User map for O(1) lookups per member.
• Error handling: Log getCachedUsers failures rather than silently discarding the error; gracefully degrades by displaying raw Slack IDs when the cache call fails.

// Before: N API calls inside loop
for _, uid := range memberIDs {
    user, err := api.GetUserInfo(uid)  // one round-trip per member
    ...
}

// After: 1 cached call + map lookup
cachedUsers, err := getCachedUsers(api)  // served from cache after first call
userByID := buildUserMap(cachedUsers)
for _, uid := range memberIDs {
    u, found := userByID[uid]  // O(1)
    ...
}

---

🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.